26 Sustainable Wireless Sensor Networks Table Battery life estimation for a senor node operating at 1% duty cycle Crossbow (2007) calculation is a very optimistic estimate as the entire capacity of the battery usually cannot be completely used up depending on the voltage drop Additionally, it is also worth mentioning that the sensors and electronics of a wireless sensor node could be far smaller than cm3 Review of Energy Harvesting Technologies for Sustainable Wireless Sensor Network 27 Hence in this case, the battery would dominate the system space usage Clearly, a lifetime of half a year for the electronic device to operate is far from sufficient because the duration of the device’s operation could last for several years This implies that the battery supply of the electronic device has to be regularly maintained The need to develop alternative method for powering the wireless sensor and actuator nodes is acute Hence the research direction is targeted to resolve the energy supply problems faced by the energy hungry wireless sensor nodes 5.2 Limitation of Power Sources for Sensor Nodes Like any other electronic devices, the sensor nodes in the WSN need to be powered by energy sources in order to operate If a wired power cable is used, many of the advantages such as self-autonomous and mobility of the sensor nodes enabled by the wireless communications are sacrificed In many applications, a power cable is not a preferable option to power the sensor nodes knowing the advantages of wireless option Instead, there are many types of portable energy systems listed in Figure.5 that are suitable for powering sensor nodes in the wireless sensor networks Among these energy systems or sources, the rechargeable/alkaline battery is one of the most popular method so far Although batteries have been widely used in powering sensor nodes in WSN presently, the problem is that the energy density of batteries are limited and they may not be able to sustain the operation of the sensor nodes for a long period of time In many application scenarios, the lifetime of the sensor node typically ranges from two to ten years depending on the requirement of the specific application Take for the case of deploying sensor nodes on the ice mountain to detect the thickness level of the ice on the mountain, it will take years for the melting process to be measurable Hence the lifetime of the sensor nodes must be to last for several years before they go into idle state If that is the case, the lifetime of one or several sensor nodes, depending on the size of the WSN, would affect the performance of the WSN Fig General types of portable energy systems Supercapacitor, in short supercap, is another electrochemical energy system other than batteries that has been gaining its presence in powering the wireless sensor nodes There are 28 Sustainable Wireless Sensor Networks Fig Ragone plot for comparing the energy storage technologies and their power density versus energy density characteristics Tester (2005) several reasons for this phenomenon to occur One reason is that supercapacitor is very scalable and its performance scales well with its size and weight Another reason is that supercap has many desirable characteristics that favour the operations of the sensor nodes such as high power density, rapid charging times, high cycling stability, temperature stability, low equivalent series resistance (ESR) and very low leakage current Referring to the Ragone plot Tester (2005) shown in Figure.6 which consolidates various energy storage technologies and compare their power density and energy density characteristics, it can be identified that supercapacitor has much higher peak power density than the other energy storage devices like batteries and fuel cells This means that supercap can deliver more electrical power than batteries and fuel cells within a short time As shown in Figure.6, the peak power densities of supercapacitors are well above 1000 W/kg level whereas the power densities of all types of batteries are in the range of 60 W/kg to 200 W/kg and fuel cells are well below 100 W/kg Hence for burst power operation, supercapacitors are better choice than batteries and fuel cells The only major drawback of supercap is that it has very low energy density as compared to the rest of the energy storage devices Batteries and fuel cells have much higher energy storage capacities than the supercapacitors, therefore they are more suitable for those energy-hungry sensor nodes that need to operate for a long time The electrical characteristics of a battery define how it would perform in the circuit and the physical properties of the battery have a large impact on the overall size and weight Review of Energy Harvesting Technologies for Sustainable Wireless Sensor Network 29 of the sensor node Batteries convert stored chemical energy directly into electrical energy They are generally classified into two groups namely 1) single-use/primary and 2) rechargeable/secondary batteries The distinction between the two groups is based on the nature of the chemical reactions Primary batteries are discarded when sufficient electrical energy can no longer be obtained from them Secondary batteries, on the other hand, convert chemical energy into electrical energy by chemical reactions that are essentially reversible Thus, by passing the electrical current in the reverse direction to that during discharge, the chemicals are restored to their original state and the batteries are restored to full charge again Some important parameters of the batteries that help to determine the performances of the battery are listed as follows: • Energy density by weight (Wh/kg) and volume (Wh/cm3 ) determines how much energy a battery contains in comparison to its weight and volume respectively • Power density by weight (W/kg) determines the specific power available per use • Self-charge rate determines the shelf life of a battery • Cost of battery The performances of the wireless sensor nodes meshed together in a network form are largely constrained by some limitations in the electrochemical type of energy system One significant limitation is the limited energy storage capabilities of the batteries or supercapacitors The energy stored in the storage elements would definitely be drained off by the connected loads after some time If this is the case, the distance range and data transmission frequency of the communication device in the sensor nodes are highly dependent upon the available electrical power supply and the electrical energy stored in the storage elements Usually, the wireless sensor networks are preferred to be left unattended once deployed in inaccessible environments where maintenance would be inconvenient or impossible, therefore replacement of the batteries in the wireless sensor nodes is out of the question The lifetime of the wireless sensor network is therefore determined by the characteristics of the batteries used In order to overcome the energy constraint of the WSN due to the energy hungry sensor nodes and the limited energy density of the storage elements, some solutions have been proposed in the next section The proposed solutions are suggested to extend and sustain the operation of the WSNs Proposed Solutions for WSN problems Often WSNs are deployed in regions that are difficult to access and so the sensor nodes should not require any maintenance at all under ideal condition They must be energetically autonomous and independent This implies that once the batteries/supercapacitors are installed for the sensor nodes, they not need to be replaced or recharged for a long period of time and really operate in an autonomous manner for life-long operation In many application scenarios, the lifetime of the sensor node typically ranges from two to ten years depending on the requirement of the specific application For that, the stringent condition imposes drastic constraints on the power consumption of the sensor node Take for an example, a single 1.5 V good AA alkaline battery is used to power a wireless sensor node for two to ten years, it can be roughly estimated that the average power consumption of the sensor node ranges from 250 µW to 50 µW Given that today’s commercially available low power radio transceivers typically consume several tens of milliwatts, keeping the transceiver constantly active is clearly impossible Several possible solutions to address these problems related to 30 Sustainable Wireless Sensor Networks powering the emerging wireless technologies have been suggested in the below list and these solutions will be further elaborated in the following sections • Improve the performance of the finite power sources for e.g by increasing the energy density of the power sources • Reduce the power consumption at different levels of the sensor nodes hierarchy i.e signal processing algorithms, operating system, network protocols and integrated circuits • Develop energy harvesting techniques that enable a sensor node to generate its own power by harvesting energy from the ambient 6.1 Improvements on Finite Power Sources Research to increase the energy storage density of both rechargeable and primary batteries has been conducted for many years and continues to receive substantial focus Blomgren (2002) The past few years have also seen many efforts to miniaturize fuel cells which promise several times the energy density of batteries While these technologies promise to extend the lifetime of wireless sensor nodes, they cannot extend their lifetime indefinitely Other than that, there are many disadvantages such as risk of fire, short shelf life of typically 2-3 years, limited energy density, low power density, etc in the existing rechargeable or alkaline batteries that are not only impacting on the operation of the sensor nodes but also causing problems to the environmental conditions 6.2 Reduce Power Consumption of Sensor Nodes Low power consumption by each individual sensor node is paramount to enable a long operating lifetime for a wireless sensor network A long sensor node lifetime under diverse operating conditions demands power-aware system design In a power-aware design, the energy consumption of the sensor node at different levels of the system hierarchy, including the signal processing algorithms, operating system, network protocols and even the integrated circuits themselves have to be considered Computation and communication are partitioned and balanced for minimum energy consumption This is facilitated by low duty cycle operation typically of the order of 0.1 % to % (most of the time the sensor nodes are sleeping), local signal processing, multi-hop networking among sensor nodes can also be introduced to reduce the communication link range for each node in the sensor network Since the loss in the communication path increases with the communication range, this reduction in the nodes linkage range would result in massive reductions in power requirements Compared with characteristics of conventional long-range wireless systems, the reduced link range and data bandwidth yield a significant link budget advantage for typical wireless sensor applications However, the severely limited energy sources (compact battery systems) for wireless sensor nodes create profound design challenges 6.3 Proposed Sustainable Power Source for WSN The wireless sensor node harvests its own power to sustain its operation instead of relying on finite energy sources such as alkaline/rechargeable batteries This is an alternative energy system for the WSN The idea is that a node would convert renewable energy abundantly available in the environment into electrical energy using various conversion schemes and materials for use by the sensor nodes This method is also known as "energy harvesting" because the node is harvesting or scavenging unused freely available ambient energy Energy harvesting is a very attractive option for powering the sensor nodes because the lifetime of the nodes would Review of Energy Harvesting Technologies for Sustainable Wireless Sensor Network 31 only be limited by failure of theirs own components However, it is potentially the most difficult method to exploit because the renewable energy sources are made up of different forms of ambient energy and therefore there is no one solution that would fit all of applications However, this option would be able to extend the lifetime of the sensor node to a larger extent compared to the other two possibilities i.e improvements on the existing finite energy sources and reduce the power consumption of sensor nodes Overview of Energy Harvesting Energy harvesting is a technique that capture, harvest or scavenge unused ambient energy (such as vibrational, thermal, wind, solar, etc.) and convert the harvested energy into usable electrical energy which is stored and used for performing sensing or actuation The harvested energy is generally very small (of the order of mJ) as compared to those large-scale energy harvesting using renewable energy sources such as solar farms and wind farms Unlike the large-scale power stations which are fixed at a given location, the small-scale energy sources are portable and readily available for usage Energy harvested from the ambient are used to power small autonomous sensors that are deployed in remote locations for sensing or even to endure long-term exposure to hostile environments The operations of these small autonomous sensors are often restricted by the reliance on battery energy Hence the driving force behind the search for energy harvesting technique is the desire to power wireless sensor networks and mobile devices for extended operation with the supplement of the energy storage elements if not completely eliminating the storage elements such as batteries 7.1 Concept of Energy Harvesting Energy harvesting systems generally consist of: energy collection elements, conversion hardware and power conditioning and storage devices as shown in Figure.7 Power output per unit mass or volume i.e power/energy density is a key performance unit for the collection elements The harvested power must be converted to electricity and conditioned to an appropriate form for either charging the system batteries or powering the connected load directly Load impedance matching between the energy collectors/energy sources and storage elements/connected to the load is necessary to maximize the usage of the scavenged energy Appropriate electronic circuitry for power conditioning and load impedance matching may be available commercially or may require custom design and fabrication Various scavengable energy sources, excluding the biological type, that can be converted into electrical energy for use by low power electronic devices are shown in Figure.7 Our environment is full of waste and unused ambient energy and these energy sources like solar, wind, vibration, ocean wave, ambient radio frequency waves, etc are ample and readily available in the environment Since these renewable energy sources are already available, it is not necessary to deliberately expend efforts to create these energy sources like the example of burning the non-renewable fossil fuels to create steam which in turn would cause the steam turbine to rotate to create electrical energy Unlike fossil fuels which are exhaustible, the environmental energies are renewable and sustainable for almost infinite long period The energy harvesting process can be easily accomplished As long as the conversion hardware are chosen correctly in relation to the energy sources, the environmental energy can then be harvested and converted into electrical energy The energy conversion hardware are designed in different forms to harvest various types of renewable energies Take for an example, the material of the photovoltaic cell in the solar panel is doped in such a way that when the solar radiation is absorbed by the cell, the solar energy from the sun would be harvested and converted into electrical 32 Sustainable Wireless Sensor Networks Fig Energy sources and respective transducers to power autonomous sensor nodes Adapted from Thomas (2006) with additional power sources energy The whole energy harvesting process involves energy conversion hardware that converts the environmental energy into electrical energy, electrical energy conditioning by the power management circuit and then store in energy storage elements and finally supply to the electrical load 7.2 Benefits of Energy Harvesting Energy harvesting provides numerous benefits to the end user and some of the major benefits about EH suitable for WSN are stated and elaborated in the following list Energy harvesting solutions can: Reduce the dependency on battery power With the advancement of microelectronics technology, the power consumption of the sensor nodes are getting lesser and lesser, hence harvested ambient/environmental energy may be sufficient to eliminate battery completely Reduce installation cost Self-powered wireless sensor nodes not require power cables wiring and conduits, hence they are very easy to install and they also reduce the heavy installation cost Reduce maintenance cost Energy harvesting allows for the sensor nodes to function unattended once deployed and eliminates service visits to replace batteries Provide sensing and actuation capabilities in hard-to-access hazardous environments on a continuous basis Provide long-term solutions A reliable self-powered sensor node will remain functional virtually as long as the ambient energy is available Self-powered sensor nodes are perfectly suited for long-term applications looking at decades of monitoring Review of Energy Harvesting Technologies for Sustainable Wireless Sensor Network 33 Reduce environmental impact Energy harvesting can eliminate the need for millions on batteries and energy costs of battery replacements 7.3 Various Energy Harvesting Techniques In both academic research works and industry applications, there are many research and development works being carried out on harnessing large-scale energy from various renewable energy sources such as solar, wind and water/hydro NREL (2010) Little attention has been paid to small-scale energy harvesting methods and devices in the past as there are hardly any need Having said that, it does not mean that there is no research activity being conducted on small-scale energy harvesting In fact, there are quite a significant amount of research works recorded in the literature that discuss about scavenging or harvesting small-scale environmental energy for low powered mobile electronic devices especially wireless sensor nodes Figure.8 shows various types of ambient energy forms suitable for energy harvesting along with examples of the energy sources The energy types are thermal energy, radiant energy and mechanical energy Fig Types of ambient energy sources suitable for energy harvesting Some energy harvesting research prototypes for harvesting various energy sources have been discussed A substantial piece of the research work done by Roundy et al in Roundy et al (2004) describes the extraction of energy from kinetic motion Roundy gave a comprehensive examination on vibration energy scavenging for wireless sensor network There are other vibration based energy harvesting research works being reported for instances piezoelectric generators in shoes Schenck et al (2001), wearable electronic textiles Emdison et al (2002) and electromagnetic vibration-based microgenerator devices for intelligent sensor systems Glynne et al (2004) In the research area of thermal energy harvesting, both Stevens Stevens (1999) and Lawrence et al Lawrence et al (2002) consider the system design aspects for thermal energy scavenging via thermoelectric conversion that exploits the natural temperature difference between the ground and air Similarly, Leonov et al Leonov et al (2007) have considered thermal energy harvesting through thermoelectric power generation from body heat to power 34 Sustainable Wireless Sensor Networks wireless sensor nodes Research on small-scale wind energy harvesting have also been performed by several group of researchers like Weimer et al Weimer et al (2006), Myers et al Myers et al (2007) and the author himself Tan et al (2007) and Ang et al (2007) Heliomote is a sensor node prototype developed by Aman Kansal et al Raghunathan et al (2005) that utilizes solar energy harvesting to supplement batteries to power the wireless embedded systems 7.4 Comparison of Energy Harvesting Sources To make the sensor node truly autonomous and self-sustainable in the WSN, the energy consumption of the sensor node must be entirely scavenged from the environment The choice of the energy harvesting technique is crucial Numerous studies and experiments have been conducted to investigate the levels of energy that could be harnessed from the ambient environment A compilation of various power densities derived from various energy harvesting sources has been listed in Table.3 Energy Source Solar (direct sunlight) Solar (illuminated office) Thermoelectric Performance (Power Density) 100 mW/cm3 100 µW/cm3 a) 60µW/cm2 5o C gradient at Blood Pressure 0.93W 100mmHg at Proposed Ambient airflow Harvester Vibrational MicroGenerators 177 µW/cm3 b) 135 µW/cm2 at 10o C gradient Piezoelectric Push Buttons µW/cm3 (human motion-Hz) 800µW/cm3 (machines-kHz) 50 µJ/N Notes Common polycrystalline solar cells are 16 %-17 % efficient, while standard mono-crystalline cells approach 20 % Typical efficiency of thermoelectric generators are ≤ 1% for ∆Thttp://www.tinyos.net< M Tubaishat & S Madria (20 03) Sensor networks: an overview, IEEE Potentials, vol .22 , pp .20 23, 20 03 A Sinha